More recently, it has been shown that a different form of shape-invariant soliton can be excited in a passive optical cavity. In addition to the dispersion-nonlinearity balance, cavity solitons (CSs) are subject to dissipation as energy is lost at each cavity round-trip due to out-coupling from the cavity. In this case, the energy loss is exactly compensated for by an external CW background field that is continuously injected into the cavity via a coupler. Because the soliton state co-exists with the CW background at all times, the shape of CSs somewhat differs from that of traditional optical solitons and does not decay to zero at large times.
Moreover, several cavity solitons can simultaneously circulate in the cavity, with almost arbitrary separations, and they can be individually addressed - turned on and off at will. This addressability has been widely studied in various contexts and been shown to lead to very rich dynamics, making CSs ideal objects for research in nonlinear optics. Originally studied in the 90’s in spatial cavities, CSs have recently attracted renewed interest in the time domain and their dynamics have been suggested to underpin specific regimes of frequency combs in micro-resonators. The interest of CSs is not restricted to fundamental nonlinear science, but they actually can have real-world applications. Perhaps the most striking characteristics of CSs is their ability to store data optically without resorting to conversion into the electrical domain, making them attractive candidates for the realization of optical buffers.
One technique to excite cavity solitons is to use an external mode-locked laser to “seed” the CSs. Although this approach is particularly appealing for its conceptual simplicity, one major problem so far has been the capacity to directly turn off existing CSs (erasure). Until now. In their latest research, Jang et al. propose and demonstrate a new scheme to individually address multiple solitons that are stored in a fiber cavity, realizing both selective excitation and erasure of CSs. Their approach builds on the fact that, by imposing a phase modulation on top of the driving CW field, soliton states can be trapped at the precise time slot where the phase gradient is minimized. By abruptly altering over a very brief moment of time the phase modulation imposed onto the driving field, they show how CSs (or bits) can be “written” at empty time slots. Even perhaps more remarkably, applying a similar perturbation during the time slots where the CSs are trapped results in sudden collapse of the CSs to the CW state, effectively erasing the original bit. This means that bit patterns can now be stored and modified arbitrarily in the optical cavity. The missing piece to the usability of CSs in optical buffering operation has finally been found.
Of course the main drawback of the technique is the relatively high CW power required to sustain CSs. As a matter of fact, CSs are nonlinear objects and can only exist if the nonlinearity of the system is sufficient. The experiments of Jang et al. should therefore be looked at as a proof-of-concept rather than an effective testbed of cavity solitons for optical memories. But with the recent progress in the development of highly nonlinear waveguides based on novel classes of glass materials, as well as high-Q monolithic micro-resonators, one can anticipate in the future not only to reduce the size of the buffer cavity but also to operate it at a much lower power consumption, a crucial step if it is ever going to replace existing solutions.
You must log in to add comments.